This is the mail archive of the binutils@sourceware.org mailing list for the binutils project.
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |
Other format: | [Raw text] |
On Thu, 7 Nov 2013, Roland McGrath wrote: > I don't know why the call was put where it was (the before_allocation hook) > rather than where I moved it (the place where the __ehdr_start symbol is > created) and the latter seems the far more obvious thing to do, so perhaps > there is some reason the former is better that I don't understand. Unfortunately this regresses the MIPS tests that I added back when making the original fix -- for the very reason the fix was placed where it was (please see http://sourceware.org/ml/binutils/2013-04/msg00260.html and http://sourceware.org/bugzilla/show_bug.cgi?id=15365 for further reference): mips-linux-gnu-ld: not enough GOT space for local GOT entries mips-linux-gnu-ld: BFD 2.24.51.20131107 internal error, aborting at .../bfd/elfxx-mips.c line 10042 in _bfd_mips_elf_relocate_section mips-linux-gnu-ld: Please report this bug. FAIL: MIPS magic __ehdr_start symbol test 1 (o32) and likewise: FAIL: MIPS magic __ehdr_start symbol test 2 (o32) FAIL: MIPS magic __ehdr_start symbol test 1 (n32) FAIL: MIPS magic __ehdr_start symbol test 1 (n64) Please try finding a better fix that can decide whether __ehdr_start can or cannot be defined no later in the link flow than where I put my change, i.e. before bfd_elf_size_dynamic_sections has been called. All symbols and their bindings have to be known by the time that function is called, you cannot add any symbols later on. > I have tested this only for {x86_64,i386}-{linux-gnu,nacl} targets. > I am not set up to test any mips targets. Setting up testing binutils e.g. for mips-linux-gnu should be fairly straightforward (unlike with other pieces of the toolchain), e.g.: $ cd path/to/binutils $ mkdir obj $ cd obj $ ../configure --target=mips-linux-gnu $ make $ make -k -i check And in any case for changes touching generic pieces of binutils I recommend regression-testing across a reasonable number of targets, e.g. I've been using a script that Alan Modra posted here a while ago (attached). I have modified it a bit to add some more MIPS targets; please feel free to tweak the list to suit your needs (and set --build appropriately). Maciej
Attachment:
binutils-test.sh
Description: Bourne shell script
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |